INDEX

Explanations

standing or waiting nearby

The neuron fires on bits of direct speech or reported‐speech-style narrative—dialogue tokens (e.g. “Yo: Quiero…,” question marks, colons) and reporting verbs (told, said, asked) that mark personal anecdotes or conversation excerpts.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

in

-1.88

now

-1.63

 every

-1.52

it

-1.45

 kafe

-1.41

of

-1.38

性は

-1.38

 more

-1.38

 pasien

-1.37

you

-1.36

POSITIVE LOGITS

 ندارد

1.55

 oídos

1.55

 نداشته

1.51

 بدون

1.50

 masculinos

1.45

biendo

1.39

 vegetación

1.35

ющихся

1.32

 suaves

1.30

without

1.28

Activations Density 0.097%